Updated on • 28 Oct 2024
Ground Sample Distance And How It Impacts Drone Surveys
A guide to GSD for drone surveying: What it is, why it is important, how it can impact the quality of your maps, and why flight altitude and a good quality camera matter.
A guide to GSD (Ground Sample Distance) for drone surveying: What it is and why it is important;
Factors that impact GSD, such as flight altitude and your camera specifications;
How to calculate GSD, using online or manual methods;
Sample datasets from the DJI M300 RTK-P1.
If you are using drones for surveying, then understanding GSD (Ground Sample Distance) is crucial.
Put simply, your GSD impacts the accuracy of your surveys. Without it, you risk collecting inaccurate data or having a map that isn’t useful.
This guide will teach you more about GSD, how to calculate it, and why flight altitude and a good quality camera matter.
What Is GSD And Why Does It Matter For Drone Surveys?
GSD is defined as the length (in inches, centimetres, or millimetres) between the centres of two consecutive pixels on your map.
Or, to think of it another way, GSD can be considered as the length of one pixel in your map.
So, if a drone achieves a GSD of 5 cm/px, that equates to one pixel on your digital map corresponding to 5 cm in reality.
Therefore, the smaller the GSD, the higher the accuracy.
Take the below images as an example: The orthomosaic with a GSD of 5 cm (left) is far more detailed that the one on the right with a GSD of 30 cm.
This is because the bigger the value of the image GSD, the lower the spatial resolution of the image and the less visible the details.
With this in mind, GSD is an important calculation for both aerial photography and photogrammetry, which is a commonly used technique for creating 3D topographic maps.
Understanding GSD as a centimetre, inch or millimeter per pixel relationship makes it easier to calculate the size of the ground and features captured in your drone image.
Not only this, but knowing the size of each pixel is necessary to grasp the full scale of your map and make decisions based on clear information.
An error of a centimetre or less may seem minor, but this mistake over hundreds of thousands of pixels, will create a serious mismatch between your map and reality, making measurements near impossible.
To play it safe, land surveyors always use the lowest possible value when calculating GSD.
GSD - Flight Altitude And Your Camera
Two key factors impact your GSD: The quality of the camera - particularly focal length and camera resolution - and flight altitude.
As a rule of thumb, the lower the altitude, the lower the GSD.
Take the DJI Phantom 4 RTK, for example. This table shows how the flight altitude impacts the GSD, ie the lower the flight, the lower the GSD.
Flight Altitude | GSD |
10 metres | 0.27 cm/px |
30 metres | 0.82 cm/px |
50 metres | 1.37 cm/px |
80 metres | 2.19 cm/px |
100 metres | 2.74 cm/px |
120 metres | 3.29 cm/px |
As well as flight altitude, a higher resolution camera further improves the GSD.
For instance, look what happens when the P4 RTK - with its 20MP sensor - goes against the DJI M300 RTK and P1 camera, with its 45MP full-frame sensor:
Altitude | P4 RTK | M300-P1 |
50m | 1.37 cm GSD | 0.63 cm GSD |
80m | 2.19 cm GSD | 1.18 cm GSD |
As you can see, the P1 achieves a lower GSD, thanks to its high-res camera. And at 50m, it actually achieves a hugely impressive sub 1cm GSD.
For the record, the below images show how a sub-centimetre GSD can help to build highly-detailed and crystal-clear maps.
The orange circle highlights a section of this orthomosaic which was captured from 50m altitude...
...and the resolution is so high that when you zoom in to that section, you can clearly count the number of holes in the bricks, or identify specific toys in the sandpit.
This level of detail is vastly important for projects requiring intricate details or measurements.
Having a higher-resolution camera also increases mapping efficiency, as it enables the drone to be flown higher - covering more ground in the process - while still capturing highly-detailed data.
As a case in point, we'll explore this theory by once again using DJI's two photogrammetry solutions: the Phantom 4 RTK (below, left) and the M300 RTK and P1 camera.
This table shows that the P1 - with its three interchangeable lenses - can be flown at a higher altitude than the P4 RTK while obtaining the equivalent GSD.
M300 RTK-P1 | P4 RTK |
24 mm, GSD = H/55; 35 mm, GSD = H/80; 50 mm, GSD = H/120. | H/36.5 |
It's not just the P1's higher-resolution which is driving the efficiency gains.
It's also the P1's pixel size, which is 1.8 times that of the P4 RTK...
....and its far larger sensor, which is seven times the size of the P4 RTK.
Larger sensors capture more light in less time: They enable a faster shutter speed to achieve sharp, well-exposed results, which improves the accuracy.
And this graphic demonstrates how all of these factors combine to make the P1 a more efficient solution. During a 15 minute flight, and with the same GSD, the M300 RTK and P1 can covered more than four times the area of the P4 RTK.
However, this is not to say that the P4 RTK is a bad drone - far from it! Rather, it demonstrates how altitude and the quality of your camera can impact GSD.
What GSD accuracy level do you need?
With this in mind, what level of GSD accuracy do you actually need? In truth, there is no right answer: It all depends on your mission.
But remember, maps comprising a higher GSD will be less accurate than ones with a smaller GSD. However, there are knock-on impacts to consider, and there are trade offs, as the table below shows:
Lower GSD | Higher GSD |
Higher accuracy | Lower accuracy |
Lower Flight Altitude | Higher flight altitude |
Longer Flight Time | Shorter flight time |
More data captured to process | Not as much data to process |
By a general rule of thumb, if you need deeply intricate measurements or a detailed reconstruction, then a lower GSD is recommended. But this does generally mean a longer flight, as less area is being covered.
On the other side of the coin, if you are conducting a large area survey that does not require highly detailed results, then upping the flight altitude and increasing the GSD is a worthwhile consideration, especially as it can reduce the acquisition time.
The below example demonstrates this.
The digger is located on our map - or chess board, as a digital photo is made up of lots of individual squares.
If we simply wanted to tell someone where the digger was, we could give any number of coordinates, such as G3 or F2.
But if we needed to calculate the width of loader, the requirement parameters would change, and more precise information would be needed. Currently, the loader is spread over two squares - D2 and E2 - which is too vague for precise calculations.
Therefore, the individual grids would need to be smaller to give a more accurate measurement. Or, to put it another way, the GSD would need to be reduced.
Essentially, choosing the right GSD will be the one that allows you to take detailed images while still flying high enough to avoid an excessive photo count.
Too high a GSD, and you’ll be left with blurry images that don’t tell you anything.
Go too low, and your survey will take up extra GB and possibly take longer than expected to complete.
This is why a solution like the P1 provides the perfect balance: Ramping up efficiency by enabling greater flight heights while collecting highly-detailed data.
How To Calculate GSD
There are two options to help you calculate GSD: Using an online tool, or doing it manually.
The first choice is simple, and there are plenty of tools to help you, such as our GSD calculator tool with presets for the latest drones and sensors.
In this tool, you enter the drone - either preset (containing most of the industry big hitters) or custom - and flight altitude, and this will provide you the GSD for the flight.
Or, you can calculate it yourself.
For this, you’ll need to know the sensor height and width, and image height and width on your drone, as well as both the focal length and flight height.
Each of these stats should be available on your drone.
You can then plug each number into two basic formulas, one for the GSD height and one for the GSD width.
GSDh= flight height x sensor height / focal length x image height
GSDw= flight height x sensor width / focal length x image width
In this case, use the GSD value which gives you the least accuracy to ensure you are using the worst-case scenario
Conclusion
The use of drones for surveying has grown exponentially over recent years. And such is the accessibility of the technology, a mapping mission can be conducted quickly and easily with a pre-planned and automated flight route.
That being said, there are factors that you need to take into consideration to ensure you collect the best data for your needs: GSD is one of these factors.
Understanding Ground Sample Distance is vital to building detailed and accurate maps: Whether that's for visual representations of a site or for conducting intricate measurements.
Knowing the capabilities of your camera, relative to flight altitude, is important.
The DJI surveying ecosystem does not disappoint in this department, and the Phantom 4 RTK is a great low-altitude, all-in-one mapping solution, especially useful to those entering the industry.
But the all-powerful M300 RTK-P1 alternative presents a highly capable option which can truly transform surveying workflows, and can compete with fixed-wing aircraft for large-area drone mapping.
written by
James Willoughby
James joined heliguy™ in 2018 following a 13-year stint in print and online journalism, having worked on regional and weekly newspaper titles. He is responsible for spearheading heliguy™'s content strategy and social media delivery. James collaborates with DJI Enterprise's European marketing team to coordinate and produce case studies and helps organise events and webinars.